# Walrus > Walrus is the verifiable data layer for high-stakes industries. Built on Sui, it provides always-available, programmable, and cryptographically authentic decentralized storage. *Complete documentation content below* # Verifiable Data Platform for AI | Walrus > Walrus is the Verifiable Data Platform for AI and onchain finance, powering Walrus Memory, the portable memory layer for AI agents. Source: https://walrus.xyz/ The foundation to store and manage your data — **persistent**, **portable**, and **under your control** across apps, providers, and agents. [Start building](https://docs.wal.app/) Introducing\ Walrus Memory ------------- **Portable memory layer** for **AI agents** [Learn more](/products/walrus-memory/) Your Verifiable Data Platform No lock-in.\ No trust gaps.\ No compromises. --------------- For too long, developers have built around infrastructure limits. **Walrus removes them**. Build products that scale without sacrificing control, portability, or correctness. [Read the docs](https://docs.wal.app/) ## Power to the builder - Reliability - Portability - Programmability ### Reliable & Verifiable Critical data stays persistent and independently verifiable, so you can trust the systems built on top of it. ### Portable Across Systems Move data across applications, providers, and agents without migrations, rebuilds, or lock-in. ### Programmable Control Control who can access data, how it's shared, and what can be done with it. Permissions stay explicit and under your control. [Read the Whitepaper](https://docs.wal.app/walrus.pdf) Partners The future runs\ on Walrus --------- Use cases ### AI Safe data in. Reliable AI platforms out. AI is only as good as the data it’s trained on. Walrus allows AI models to authenticate and verify workloads, preventing poisoned datasets or corrupted results. [Learn more](/use-case/ai/) ### Data Markets Your data. Your terms. No middleman. Build open, verifiable data marketplaces where creators and businesses can publish data and get paid. [Learn more](/use-case/data-markets/) ### DeFi Real-time protection for the digital economy Provide users with always-on services and shield online transactions from bad actors by authenticating websites and interactions in real time. [Learn more](/use-case/defi/) The wal token The economic\ heart of Walrus --------------- WAL is used to pay for storage and secure the network through staking. [Learn about WAL](/use/) Part of the\ Sui Stack --------- Created by **Mysten Labs**, the team behind **Sui**, Walrus is a core part of the Sui Stack. This modular platform brings execution, privacy, compute, and liquidity together so builders can focus on creating great apps, not managing complexity. [Explore on Sui](https://sui.io/) ## FAQs ### What makes data on Walrus verifiable? Data on Walrus is independently verifiable, meaning anyone can confirm it is correct and unaltered without trusting the provider storing it. Verifiability is a native property of the platform, rather than added as an afterthought, so the apps and agents built on top can trust every byte they read and write. ### Does my data stay available and persistent over time? Yes. Walrus keeps critical data persistent and continuously available across a decentralized network, with no single provider to depend on. Your data remains durable and retrievable over time, which is what enables agents to rely on persistent memory and apps to preserve records they cannot afford to lose. ### Who controls access to data stored on Walrus? You do. Walrus gives you programmable control over who can access your data, how it is shared, and what can be done with it. Permissions stay explicit and under your control, so portability never means giving up ownership or exposing data you meant to keep private. ### How do I give my AI agent persistent, verifiable memory? Use [Walrus Memory](/products/walrus-memory/), the portable memory layer for AI agents built on the Walrus Verifiable Data Platform. Agent memory, conversations, checkpoints, and reasoning, is stored as data that persists across sessions and stays independently verifiable, so your agent can prove what it remembered and trust what it reads back. ### How can I make my agent's memory portable across apps? [Walrus Memory](/products/walrus-memory/) stores agent memory as portable data that moves across apps, providers, and workflows with no migrations, rebuilds, or lock-in. Because memory lives on Walrus rather than inside a single app or vendor, any authorized app or agent can read and build on the same memory layer. ## Get your data right Step-by-step guides to get you up & running fast [Watch tutorials](https://www.youtube.com/@Walrus-Protocol) Learn how to get started on Walrus [Start building](https://docs.wal.app/) Get the support you need to build bigger [Apply for grants & RFPs](/rfp/) --- # What is Walrus? | Verifiable Data Built on Sui > Created by Mysten Labs, the team behind Sui, Walrus removes the infrastructure limits holding developers back. Always-on, verified, and programmable at scale. Source: https://walrus.xyz/about/ About Walrus Speed and trust matters.\ Walrus has your back. [Read the docs ](https://docs.wal.app/)[Learn about WAL](/wal-token/) Secure Reliable Fast Your data layer\ is holding you back ------------------- Growth requires infrastructure you can trust. But today's data layers are haunted by downtime, slow performance, and inflexibility that turns ambitious ideas into endless compromises or prevents builders from shipping at all. For high-stakes apps, unreliable infrastructure isn't just frustrating, it's a threat that can derail launches, limit possibilities, and create real financial consequences. ### Walrus gives builders the tools they need Created by Mysten Labs, the team behind [Sui](https://sui.io), Walrus is the solution for builders seeking to protect, accelerate, and activate data to achieve growth. The result: a platform that keeps data available and secure, delivers cryptographic verification quickly, and makes data programmable through smart contracts. 356TB active data stored 200+ ecosystem partners \~800ms downloads for small files < 1MB How it works ## Your data deserves better Traditional storage just sits there.\ Walrus is a new way of thinking about data. ### Decentralized by design Data is spread across decentralized storage nodes, resilient to failure even if 2/3rd of nodes fail ### Fast reads and writes A global network reconstructs and delivers your data in milliseconds ### Publicly verifiable Onchain proofs let you verify your data independently, anytime ### Unlock new possibilities Power credible AI systems, turn data into revenue, and make DeFi dependable Walrus Foundation ## Investing in trustedbuilders Walrus Foundation is a not-for-profit organization dedicated to the promotion and growth of the Walrus ecosystem. The organization supports builders through grants and RFPs, helping developers tackle the biggest challenges in high-stakes data. [Learn more](/rfp/) ## Time to build without compromises Learn how to get started on Walrus [Read the docs](https://docs.wal.app/) Discover how WAL powers the future of onchain data [Learn about WAL](/wal-token/) --- # How Alkimi Made Every Ad Click Verifiable | Walrus > Alkimi processes 25M daily ad impressions on Walrus, replacing centralized storage with cryptographic proof and a fully auditable ad exchange at scale. Source: https://walrus.xyz/case-study/alkimi/ Case Study Discover how they're using Walrus to make every click verifiable. Every day, advertisers spend millions on ads that may never be seen by real people. Brands can't verify what they're paying for, consumers are tracked across every click, and the entire system runs on complex infrastructure riddled with middlemen and extra fees. Walrus gave Alkimi, a blockchain-based ad exchange, the data foundation needed to solve this: a tamper-proof, auditable system where every impression is provably real, every click is verifiable, and “just trust us” is replaced with cryptographic proof. 25M daily ad impressions 3B+ transactions 15GB of data daily ## Can your data layer keep up? Alkimi started out with a simple challenge: get a data layer that could keep up with the fast-paced world of advertising, where billions of clicks, impressions, and transactions flow through the system daily. They relied on validators for consensus and recording key transaction metadata, but storing and reproducing the full volume of execution data onchain was too slow and expensive. Their previous solution involved centralized storage, leaving their customers dependent on a single provider's uptime and payment processing for data accessibility, which also undermined their goal of a fully decentralized system. ## Proving every click Walrus and the [Sui Stack](https://blog.sui.io/alkimi-onchain-advertising-sui-stack/) gave Alkimi exactly what they needed: the infrastructure necessary to handle massive volumes of data without compromising on speed, cost, or auditability. Instead of relying on a single provider, Walrus's replicated storage spreads data across a distributed network—lowering costs per gigabyte while making everything more durable and available. Now Alkimi can keep every impression, every click, every transaction auditable and verifiable. > “The Walrus integration was seamless. Clear, well-organised documentation meant our team spent less time troubleshooting and more time building.” Alkimi Engineering Team ## Delivering advertising at scale With Walrus as its data foundation, Alkimi can deliver verifiable, near-real-time reconciliation and settlement between advertisers, agencies, and publishers. The platform maintains full-fidelity audit trails that satisfy compliance requirements and give partners unprecedented assurance into campaign performance, which is critical for informing future spending. Plus, they've reduced all the risks of a centralized provider, improving overall reliability when it matters most. ## Start building with verifiable data Learn how to get started on Walrus [Start building](https://docs.wal.app/) --- # How Baselight Built a Permissionless Data Economy | Walrus > Baselight integrated Walrus in two weeks to build a permissionless data platform where datasets are instantly queryable, monetizable, and AI-ready. Source: https://walrus.xyz/case-study/baselight/ Case Study See how they’re using Walrus to help companies turn data into opportunity. Gaining access to the data you need to build intelligent apps shouldn’t take days of waiting for approvals from centralized gatekeepers, but that’s exactly how most data distribution works today. This outdated system slows down AI development, blocks research, and keeps valuable datasets out of reach for the teams that need them most. Walrus enables Baselight, a permissionless data platform, to turn stored files into living assets that people can activate for monetization, queries, and AI training. Instead of waiting days for approval, data becomes instantly discoverable and queryable, powering new use cases in AI and beyond. ## When scaling up means slowing down Baselight wanted to give customers the option of a decentralized data layer to avoid the tradeoffs, including a lack of control and dependence on a single entity, that come with centralized cloud providers. But that flexibility seemed to come at a cost: slow queries across different storage systems, inconsistent performance, and rising expenses when handling lots of small files like logs or AI outputs. As datasets grew and more users came onboard, these issues would only compound, making it harder to scale without sacrificing speed or reliability. The real challenge was finding decentralized infrastructure built to handle growth without falling apart. ## Activating the value of data Walrus empowered Baselight with the decentralized infrastructure it needed to scale without performance tradeoffs. In just two weeks, the team was able to integrate Walrus and offer users a data layer that handles both massive datasets and thousands of small files (logs, AI outputs, sensor data) without slowdowns. Walrus also unlocked features Baselight couldn't easily build before: programmable storage that enables automatic payments to data providers whenever their datasets get queried, and enforceable access controls that protect sensitive data. The result is a system that stays fast, flexible, and truly decentralized as it grows. > “Baselight is building the permissionless data economy, and Walrus provides the decentralized data management layer that makes it possible.” Henrique Moniz, CEO at Finisterra Labs (creator of Baselight) ## Faster data for the future of tech With Walrus, Baselight’s clients now have access to a decentralized solution with lightning fast data retrieval. This is especially critical when working with high-volume structure data, like sensor streams and AI agent outputs, that are helping shape the future of technology. ## Explore Walrus for data markets Learn how to get started on Walrus [Start building](https://docs.wal.app/) --- # Discover Bluefin's blueprint for verifiable trading | Walrus > Bluefin, a decentralized financial ecosystem that has processed over $100B in trading volume, is rebuilding the exchange from the ground up. Source: https://walrus.xyz/case-study/bluefin/ Case Study Creating fast, verifiable DeFi data with public proof — powered by Walrus. For years, DeFi has faced an impossible trade-off: speed or decentralization. Blockchains provide secure custody and settlement, but they are not optimized for millisecond-level order matching. Even sub-second block times introduce enough latency to kill market-maker risk tolerance. Bluefin, a decentralized financial ecosystem that has processed over $100B in trading volume,has rebuilt the exchange from the ground up with the introduction of Bluefin Pro, which will use Walrus to drive millisecond execution, non-custodial funds, and a fully verifiable audit trail. ## DeFi got faster-but verification got harder While offchain execution solves the speed problem, it creates a new one. Users must trust the operator to execute fairly and publish accurate results. Most offchain systems store execution history inside the operator's own infrastructure. If something goes wrong, users have no independent way to check. ## Onchain custody meets verifiable execution Bluefin Pro separates three things that were previously bundled together: custody, execution, and verification. Custody and settlement stay onchain. Order matching moves offchain into a Trusted Execution Environment powered by [Nautilus](https://www.sui.io/nautilus). And every action (open order, cancel, match) is logged to Walrus in real time. Checkpoints are published every few seconds, allowing validators to efficiently sync and verify orderbook state without replaying the entire history. Code binaries will be fully open-source. Anyone running Bluefin's matching engine can reproduce the orderbook state and produce the same attestations independently. ## How it works ### Deposit Onchain Users deposit into non-custodial Sui smart contracts. Funds stay under user control throughout. ### TEE Matching Order matching occurs inside a TEE powered by Nautilus. Every action produces a cryptographic attestation, published to Walrus in real time. ### Settle & Verify Only final balance changes go onchain. Full execution history lives on Walrus that anyone can verify. No permission needed. ## $100B in volume: now verifiable Bluefin has processed over $100B in trading volume since launching on Sui in Q4 2023 and Bluefin Pro represents a complete rebuild of that exchange on a verifiable foundation. ### User custody Funds stay in non-custodial Sui smart contracts. The TEE never holds assets. ### CEX-like Speed Sub-millisecond order matching. Fast cancels mean tighter spreads and deeper books. ### Public Verifiability Anyone can reproduce the TEE's computation from inputs and outputs on Walrus. No permission needed. ### Onchain Settlement Final balances, margins, and liquidations enforced by Sui smart contracts. ### Open Source All code binaries will be open-source and anyone running the matching engine can produce the same attestations. Bluefin demonstrates the pattern for a derivatives exchange and the same architecture applies to any system running fast offchain computation where the results carry real financial stakes and independent verification matters. Explore DeFi on Walrus [Learn more](/use-case/defi/) Learn how to get started on Walrus [Start building](https://docs.wal.app/) --- # How Inflectiv Made AI Agent Data Trustworthy | Walrus > Inflectiv uses Walrus to make AI agent data verifiable and traceable, storing over 7,000 datasets with 60% lower costs than centralized alternatives. Source: https://walrus.xyz/case-study/inflectiv/ Case Study Learn how they used Walrus to help AI agents make intelligent decisions. AI agents often fail not because the models are weak, but because the data pipelines feeding them are unverifiable. Apps building in critical industries, like DeFi and AI, need data that is traceable, yet no trust layer exists between data ingestion and agent execution. Walrus gave Inflectiv, the intelligence infrastructure for AI agents, the tool to close that gap, replacing centralized storage with cryptographically verifiable, immutable data. Now enterprises and Web3 systems can finally trust what their agents know—and prove it. 60% reduction in cost\ vs AWS S3 4,152+ datasets stored\ on Walrus < 2\ weeks integration time ## Good decisions start with good data Clean, reliable intelligence is what separates agents that make smart decisions from those that don’t. Inflectiv transforms messy, unstructured data—like PDFs, internal docs, web pages, sensor data—into structured, validated datasets that power AI agents in production systems. In order to do so successfully, they needed to guarantee data powering agents hadn’t been altered, could be traced back to its original source, and would hold up in high-stakes industries like DeFi and AI. As AI agents take on higher-stakes decisions, the ability to independently audit what data they acted on is quickly becoming an expectation for both enterprise and regulated environments. ## Store. Verify. Prove. Walrus delivered exactly that, combining decentralized storage with onchain provenance in a single, integrated stack. Every raw file ingested by Inflectiv is anchored as an immutable blob on Walrus, assigned a cryptographic blob ID, and registered as an object on Sui, creating an unbroken chain of custody from raw data all the way through to agent execution. Where centralized storage could only offer platform trust, Walrus offered something far more durable: independent, verifiable proof that the data is exactly what it claims to be. > “Walrus was surprisingly straightforward to integrate. The publisher API is clean, and pairing it with Seal for per-file encryption gave us decentralized access control that would have taken weeks to build on AWS.” Inflectiv Engineering Team ## Now verifying over 4,000 datasets Since integrating Walrus, Inflectiv has stored over 4,000 datasets. The architecture shift has been equally significant. What once relied on centralized cloud infrastructure now runs on a trust-minimized stack where every dataset version is immutable, every agent decision is traceable, and every source can be independently verified. ## Learn how Walrus powers verifiable AI Learn how to get started on Walrus [Start building](https://docs.wal.app/) --- # Walrus Ecosystem | 200+ Partners Building on Verifiable Data > From AI to DeFi, 200+ organizations are building on Walrus. Explore partners, case studies, and use cases powered by verifiable data. Source: https://walrus.xyz/ecosystem/ From AI agents to DeFi infrastructure, organizations are using Walrus to make the most of their critical data. Join the more than 200 partners already building on a foundation of secure, always-available infrastructure. From AI agents who can work across multiple data sources to wearables that make monetizing data simple, it all starts on Walrus. Case Studies [Revolutionizing traditional adtech](/case-study/alkimi/) [View Case Study](/case-study/alkimi/) [Blueprint for verifiable trading](/case-study/bluefin/) [View Case Study](/case-study/bluefin/) [Verifiable intelligence for AI agents](/case-study/inflectiv/) [View Case Study](/case-study/inflectiv/) Walrus has your back [Build now](https://docs.wal.app/) --- # Walrus Network Release Schedule > Stay up to date on upcoming Walrus network releases, upgrades, and scheduled maintenance across Testnet and Mainnet. Source: https://walrus.xyz/network-release-schedule/ Get the latest on what's shipping and when. ## Testnet The Walrus Testnet operates on the Sui Testnet and is used to test new features of Walrus before they graduate to the Mainnet. Number of shards | 1000\ Epoch duration | 1 day\ Max. number of epochs for which storage can be bought | 53 ## Mainnet Walrus Mainnet operates a production-quality storage network on the Sui Mainnet. Number of shards | 1000\ Epoch duration | 2 weeks\ Max. number of epochs for which storage can be bought | 53 [Learn more](https://docs.wal.app/docs/usage/networks) --- # Subscribe to the Walrus Newsletter > Get the latest Walrus news, ecosystem updates, and developer resources delivered straight to your inbox. Source: https://walrus.xyz/newsletter/ Newsletter Learn how Walrus is powering the next generation of AI using the Verifiable Data Platform that makes your data persistent, portable, and yours to control. --- # Portable Memory Layer for AI Agents | Walrus Memory > Portable agent memory that moves across apps and sessions. Walrus Memory enables reliable coordination and programmable access control for AI agents. Source: https://walrus.xyz/products/walrus-memory/ Walrus Memory is a portable memory layer that makes AI agents reliable across apps and sessions. Persistent, verifiable, and fully under your control. [Start building ](https://memory.walrus.xyz/)Watch video Close Connect an agent Build agents Python Build apps TypeScript Add Walrus Memory to **Claude Code**, **Cursor**, **Codex**, **Gemini CLI**, and more. Run \`curl -sL https\://memory.walrus.xyz/skills/setup\` and use the returned instructions to set up Walrus Memory in this AI client. Create multi-agent workflows with shared memory. \# pip install memwal ​ import asyncio import os from memwal import MemWal, RecallParams ​ async def main(): walrus\_memory = MemWal.create( key=os.environ\["MEMWAL\_PRIVATE\_KEY"], account\_id=os.environ\["MEMWAL\_ACCOUNT\_ID"], env="prod", namespace="demo", ) ​ await walrus\_memory.health() await walrus\_memory.remember\_and\_wait("I live in Hanoi and prefer dark mode.") ​ result = await walrus\_memory.recall(RecallParams(query="What do we know about this user?")) for memory in result.results: print(memory.text, f"(distance: {memory.distance:.3f})") ​ await walrus\_memory.close() ​ asyncio.run(main()) Add Walrus Memory directly to your application. // npm install @mysten-incubation/memwal // pnpm add @mysten-incubation/memwal // yarn add @mysten-incubation/memwal ​ import { MemWal } from "@mysten-incubation/memwal"; ​ const walrusMemory = MemWal.create({ key: "\", accountId: "\", serverUrl: "https\://your-relayer-url.com", namespace: "demo", }); ​ await walrusMemory.health(); const job = await walrusMemory.remember("I live in Hanoi and prefer dark mode."); await walrusMemory.waitForRememberJob(job.job\_id); ​ const result = await walrusMemory.recall({ query: "What do we know about this user?" }); console.log(result.results); Builders shipping\ with Walrus Memory ------------------ > “Portable memory across AI systems is a huge unlock. Engineers already bounce between OpenAI, Anthropic, and Gemini, and switching between platforms means rebuilding context from scratch. Walrus Memory is helping make persistent, portable context a foundational piece of AI infrastructure.” > “Walrus Memory is going to let our monitoring agents retain prior observations, so they can avoid reprocessing the same events. Every decision will also produce an audit trail, allowing us to verify and explain why activity was flagged.” > “As AI systems run on their own and for longer stretches, their memory infrastructure calls for stronger guarantees around verifiability, portability, and reliability. We see Walrus Memory as part of a major shift toward open and interoperable AI systems.” Without memory,\ agents hit reset ---------------- A workflow spans two agents, and neither knows what the other did. A session ends and the context goes with it. A workflow spans two agents, and neither knows what the other did. A complex job is incomplete, with no checkpoint and no recovery. State isn't shared. Agents make conflicting decisions. Memory is stale or unverifiable. Output isn't reliable in production. Something breaks. No audit trail, no way to trace what your agent acted on. ## A portable memory layer for AI agents ### Portable by design Context doesn't die when the session ends. Your agent picks up where it left off — in a different app, in a different runtime, weeks later. Your agents keep context wherever they run ### Yours to control You decide how every memory is shared, accessed, and updated. Programmable permissions keep privacy on your terms. Explicit ownership and access control ### Built for coordination In multi-step workflows, your agents share memory and stay coordinated on the same state. Memory integrity is independently verifiable. Multiple agents, one source of truth [Get started](https://memory.walrus.xyz/) ## Plugs into the stack you already have Walrus Memory ships with native support for the platforms and protocols teams are already building on. [Explore the Docs](https://docs.wal.app/walrus-memory/getting-started/what-is-walrus-memory) - Works with every major LLM out of the box - First-party plugins for OpenClaw and NemoClaw - Native MCP support. No adapters needed - Python, TypeScript, and JavaScript. Pick your language [Explore the Docs](https://docs.wal.app/walrus-memory/getting-started/what-is-walrus-memory) ## What builders are creating with Walrus Memory A portable agent memory layer doesn’t just improve your stack. It changes how your product behaves — and what users come to rely on. ### Multi-agent workflows Agents share context across tasks, tools, and time. What one agent learns can become available to the entire workflow. ### Customer support agents Pick up where the last conversation ended. Know what was tried, what failed, and what the customer actually needs. ### Notes and research apps A second brain that compounds over time. It captures ideas continuously and retrieves them naturally weeks later. ### Personal assistants Remember tone, preferences, routines, and relationships. Feel like assistants that know you, not new hires every session. [Start building](https://memory.walrus.xyz/) ## Memory that holds up in production Context travels with the workflow: programmable permissions, verifiable state, no lock-in. Built on [Walrus, the Verifiable Data Platform](https://walrus.xyz/). ### Remember Authenticate Your SDK sends a signed request so the relayer can verify ownership and access. Process The relayer analyzes and encodes your memory using embedding models optimized for retrieval. Encrypt & store Your memory is encrypted, stored on Walrus, and indexed for fast semantic search. ### Recall Authenticate Your SDK sends a signed query request so the relayer can verify access permissions. Search The relayer interprets the query, retrieves the most relevant memories, and fetches them from Walrus. Decrypt & inject The memory is decrypted and injected into the model context for downstream reasoning. Build agents\ that remember ------------- Wire memory into your agent.\ Ship without rebuilding context every session. [Start building](https://memory.walrus.xyz/) [Explore Docs ](https://docs.wal.app/walrus-memory/getting-started/what-is-walrus-memory)[Go to GitHub ](https://github.com/MystenLabs/MemWal)[Join Discord](https://discord.gg/walrusprotocol) ## FAQs ### What is Walrus Memory? Walrus Memory is a portable memory layer for AI agents. It runs on [Walrus, the Verifiable Data Platform](https://walrus.xyz/), and gives agents persistent, shared, and verifiable memory across sessions, apps, and runtimes. ### Why do AI agents lose memory between sessions? AI agents often lose memory because most agent experiences are session-based. When the chat, app, or runtime ends, the context window disappears with it. The next session starts from scratch unless the developer has built a persistent memory layer. That's the source of AI agent memory loss: memory is tied to the session, not the agent. Walrus Memory makes the memory portable instead, so your agent picks up where it left off in a different app, in a different runtime, weeks later. ### How do AI agents store memory securely? AI agents store memory by saving context they can retrieve and use later. Most teams build this themselves with vector databases, storage systems, application state, and custom access controls. It works, but often creates fragmented infrastructure with limited portability and few guarantees around ownership, access, or memory integrity. Walrus Memory keeps memories encrypted before they are stored, while ownership and delegate permissions determine who can access them. Access controls are enforced and independently verifiable, giving developers explicit control over how memory is accessed and shared without building custom infrastructure from scratch. ### How can I make my agent's memory portable across apps? Portable AI memory means decoupling memory from the app the agent runs in, so the same context can be read and written from anywhere. Walrus Memory does this by storing every memory on [Walrus — the Verifiable Data Platform](https://walrus.xyz/) — with programmable permissions that travel with it. Your agent authenticates with a signed request from the SDK and pulls the same memory whether it's running in a chat app, a workflow, or a different framework entirely. No migrations, no adapters, no lock-in. ### How do I give my AI agent persistent memory? You can give your AI agent persistent memory by installing the Walrus Memory SDK in Python or TypeScript and wire two calls into your agent: one to remember, one to recall. The SDK handles authentication, encryption, embeddings, and storage on Walrus, so you don't manage vector indexes or infrastructure yourself. With native MCP support and first-party plugins for OpenClaw and NemoClaw, your agent gets persistent memory across sessions, apps, and runtimes — without rebuilding context every time. ### How is Walrus Memory different from Redis, Postgres, or a vector DB? Redis, Postgres, and vector databases are storage primitives. Developers can use them to build agent memory, but they still need to handle things like embeddings, access control, encryption, ownership, and portability themselves. Walrus Memory is a memory layer built specifically for AI agents. Memories are encrypted, semantically indexed, stored on Walrus, and managed through ownership and delegate permissions. Developers get persistent, portable memory with programmable access controls built in. ### Who owns the memories stored in Walrus Memory? You do. Every memory is tied to a signing key you control — only requests signed by that key can read or write it. Permissions are programmable, so you decide which agents access which memories and on what terms. Nothing is shared by default, and no middle layer can read your memories. ### Where is memory stored and is it encrypted? Memory is stored on [Walrus — the Verifiable Data Platform](https://walrus.xyz/). Every memory is encrypted before storage and decrypted only when a signed request from an authorized SDK retrieves it. Integrity is independently verifiable, so you can prove a memory hasn't been tampered with. ### How do I get started with Walrus Memory? [Create a Walrus Memory account](https://memory.walrus.xyz/), generate your delegate key, install the SDK, and you’re off to the races. During the launch period, builders can get started with Walrus Memory for free. Usage limits apply, and pricing may evolve over time. Walrus Memory is a portable memory layer that helps AI agents carry context across apps and sessions without losing continuity. --- # Build the Verifiable Internet | Walrus Grants and RFPs > Walrus Foundation funds builders tackling hard problems in verifiable data. Apply for a grant or open RFP to get financial support and ecosystem backing. Source: https://walrus.xyz/rfp/ Supporting builders creating the next generation of apps with verifiable data. [View open RFPs](https://airtable.com/appoDAKpC74UOqoDa/shr1je0hfpi4LFHHx/tbliqV4teM5mxdDVp) ## Powering the verifiable future Someday, the entire internet will run on data that's verifiable, programmable, and secure. Walrus Foundation funds builders who are making this dream a reality via grants and RFPs in areas of critical need. We're looking for teams tackling the hardest problems in high-stakes systems, where trust, speed, and reliability can't be compromised. Funded by Walrus Foundation [Essential AI tools for defining styles, extracting distinct assets, and composing freely.](https://bellyseal.com) [Explore](https://bellyseal.com) [Publishing platform where creators fully own their content, distribution, and monetization opportunities.](https://inkray.xyz/) [Explore](https://inkray.xyz/) [Decentralized crowdfunding for communities, creators, and startups.](https://crowdwalrus.xyz/) [Explore](https://crowdwalrus.xyz/) [Create and deploy static sites to Walrus decentralized storage. AI-powered editor, templates, and live preview — all from your browser.](https://walgo.xyz/) [Explore](https://walgo.xyz/) Criteria ## How projects are evaluated ### Technical strength and execution Present realistic milestones and a clear development plan, and demonstrate your team's ability to build & ship. ### Creativity Display creative approaches that explore and give vision to what will be built. ### Engagement in the ecosystem Showcase active building in the ecosystem and ongoing collaboration with the Walrus community. ### Team commitment and resourcing Exhibit momentum towards long-term success, with additional resources, funding, or a clear path to sustainability. ## Grants Grants, which include both financial support and help getting started, are currently available to builders active in our ecosystem. If you know someone on the team, reach out and have them sponsor or recommend you. We love building with people we trust. New to our community? No problem.\ Here's how to get started: Say hi at events [View events](https://luma.com/user/walrus) Get active on Discord [Join Discord](https://discord.gg/walrusprotocol) ## FAQs ### Who is eligible to submit a proposal for an RFP? The Walrus Foundation’s request for proposals welcomes applications from any individual, team, organization, or entity passionate about the possibilities of verifiable data. ### How many proposals are accepted per RFP? In most cases, the Walrus Foundation accepts one proposal per request. However, there may be circumstances where multiple proposals are accepted. ### How are RFP proposals evaluated and selected? Evaluation criteria include factors such as technical feasibility, innovation, relevance to the RFP's objectives, potential impact on the Walrus community, and the applicant's track record or experience. After review, the shortlisted proposals undergo discussion and final voting by an RFP committee to determine the candidate with the highest potential to address the RFP's requirements and contribute positively to the Walrus ecosystem. ### Can I submit a proposal with others? Yes, if you are an individual and you believe the scope of work requires a team, we encourage you to partner with others. ### Is there a closing date for applications? Application review will start \~2 weeks after each RFP is opened. RFPs will remain open until a suitable team or solution is selected. ### How long does it take to receive a decision on a proposal? Due to the nature of the program, we are receiving a high number of submissions. While we aim to respond to every team, whether selected or not, our immediate focus will be on reaching out to the teams chosen to move forward. ### Is there financial support for the selected proposal? Yes, the final agreement includes funding and a budget for the winning applicant. Payment timing is dependent on agreed upon milestones that reflect meaningful progress. ### Aside from financial support, are there other forms of support available? After a team is selected, we will open a direct line of communication between the selected team and the Walrus Foundation. We encourage selected teams to communicate with us to discuss what other forms of support would be most beneficial, as these may differ from team to team. ### Who is eligible to apply for a grant? Rather than a purely form-based application process, grants are typically awarded through ongoing conversations with teams we’ve gotten to know through the community. If you’re interested, we encourage you to share what you’re building, join the Discord, come to events, and get involved — we’d love to connect and explore whether a grant could be a good fit. ### How are grants evaluated and selected? Walrus grants are milestone-driven ecosystem investments. Proposals are evaluated based on ecosystem relevance, technical feasibility, innovation, and the potential to drive meaningful adoption of Walrus. All grants are milestone-based, with success measured through clear KPIs such as developer usage, integrations, and real activity on Walrus storage. Grants typically support smaller, well-scoped deliverables or MVP-level traction experiments focused on validating demand, demonstrating early ecosystem impact, and helping projects build the momentum needed to attract larger external investment. Selected teams are reviewed internally and supported through ongoing discussions to ensure alignment on scope, milestones, and outcomes that contribute positively to the Walrus ecosystem. ## The verifiable internet needs you [View open RFPs](https://airtable.com/appoDAKpC74UOqoDa/shr1je0hfpi4LFHHx/tbliqV4teM5mxdDVp) --- # Let's talk about your agents' memory | Walrus > Tell us what you're building and we'll follow up to set up time, a demo, a deep dive, or just a conversation about agent memory. Source: https://walrus.xyz/talk-to-us/ Contact Us Walrus Memory is a portable, verifiable memory layer for AI agents, built by the ex-Meta team behind Sui and already integrated with leading models and frameworks. Come talk to us at HumanX! A few quick details, and we'll follow up. --- # Tusk Crasher | Walrus: Data Matters > Add the Walrus to your photo and show the world data matters. Tusk Crasher: the world's most verifiable photo invasion. Source: https://walrus.xyz/tusk-crasher/ Show the world data matters Upload your profile pic, get the tusked version, re-upload to X --- # Verified Data for AI Development | Walrus > Unverified training data is AI's biggest risk. Walrus makes datasets and AI outputs verifiable by default, so developers can build platforms users trust. Source: https://walrus.xyz/use-case/ai/ Your AI learned from somewhere.\ Verifiable data means you can vouch for it. AI Partners on Walrus The problem AI is eating the world, but its infrastructure remains stuck in the past. Centralized servers, unverified training data, and minimal accountability for where data comes from are all major risks. The solution Walrus makes datasets, models, and content verifiable by default, so developers can secure AI platforms from potential regulatory non-compliance, inaccurate responses, and erosion of user trust. 87% of AI projects fail before reaching production due to data quality problems ([Source](https://www.akaike.ai/resources/the-hidden-cost-of-poor-data-quality-why-your-ai-initiative-might-be-set-up-for-failure)) 60% of AI projects will be abandoned due to lack of AI-ready data ([Source](https://www.gartner.com/en/newsroom/press-releases/2025-02-26-lack-of-ai-ready-data-puts-ai-projects-at-risk)) 45% of business leaders cite concerns about data accuracy of bias as a leading barrier to scaling AI initiatives ([Source](https://www.ibm.com/think/insights/cost-of-poor-data-quality)) Don't wait for verifiable AI.\ Make it. -------- Get the latest Walrus AI product updates, right in your inbox. ## Ready to start building? [Explore the docs](https://docs.wal.app/) Opportunities The future of AI\ is verifiable ------------- What can you do with Walrus? ### Deploy autonomous AI agents Build offchain agents with onchain permissions, enabling new, automated workflows and payments. ### Create AI agents with identity & memory Give agents memory and verifiable reputations, so users can recognize, trust and interact seamlessly. ### Train models on verified, licensed content Access high-quality datasets from publishers with cryptographic proof of permissions, ensuring transparent and compliant AI training. ### Store verifiable training datasets Build pipelines with cryptographic proof of data provenance and history. ### Prove authenticity of AI outputs Link responses back to the model, prompt, and training data used. ## Unlock the AI agent revolution “With Walrus, Talus AI agents can seamlessly store, retrieve, and process data onchain—empowering developers to efficiently build, deploy, and scale AI agents.” Mike Hanono, Co-Founder and CEO at Talus “Walrus offers us a highly performant yet cost-effective solution that allows us to maintain the speed Everlyn has become known for, making this partnership a no-brainer.” Dr. Sernam Lim, Cofounder and CEO of Everlyn ## Start building provable AI [Try Walrus](https://docs.wal.app/) --- # Build Verifiable, Programmable Data Markets | Walrus > Data markets fail when data can't be trusted. Walrus gives developers verifiable provenance and programmable permissions to turn data into a tradable asset. Source: https://walrus.xyz/use-case/data-markets/ Turn data into revenue on Walrus, where every byte of data is programmable and verifiable. Data Markets partners on Walrus The problem Data markets don’t work when data can’t be trusted. Buyers won’t purchase what they can’t verify, users won’t sell their personal data without restrictions, and billions in value remains out of reach. The solution Walrus breaks the stalemate, allowing developers to build data markets with verifiable data. With access control and cryptographic verification, builders can transform the data industry by making data a securely tradable asset. $3.47B value of the global data monetization market in 2024 39% of users would share their data for monetary compensation ([Source](https://www.prnewswire.com/news-releases/survey-shows-consumers-very-willing-to-trade-personal-data-for-financial-benefits-301106196.html#:~:text=Consumers%20recognize%20their%20personal%20data,keep%20their%20personal%20data%20safe.)) Don't just watch the data economy change.\ Build it. --------- Get the latest tutorials, partner news, and network updates from the Walrus team. ## Ready to start building? [Explore the docs](https://docs.wal.app/) Opportunities Infrastructure that\ unlocks value ------------- What can you build with Walrus? ### Verifiable provenance Empower buyers with guaranteed data authenticity. ### Data monetization Enable users to turn data into real revenue streams. ### Programmability Set custom terms for data usage with Sui smart contracts. ### Granular permissions Protect user privacy with access control, allowing them to control what data is viewed and sold. ## Create trusted data markets [Try Walrus](https://docs.wal.app/) --- # Verifiable Data Infrastructure for DeFi | Walrus > Bad data in DeFi means bad pricing and lost funds. Walrus gives protocols verifiable market data, auditable execution records, and frontends that stay online. Source: https://walrus.xyz/use-case/defi/ From governance artifacts to oracle inputs and settlement data, Walrus keeps DeFi records durable, inspectable, and resilient. DeFi apps on Walrus The problem In DeFi, bad data is more than a nuisance. It directly affects pricing, execution, liquidations, and trust, yet many protocols still rely on centralized storage and hosting for critical market data, vault execution records, and even their frontends. The solution Walrus is built for systems where correctness and availability are non-negotiable, allowing teams to store, serve, and reference DeFi-critical assets in a way that stays accessible and provable. $105B Total value locked in DeFi ([Source](https://www.coindesk.com/business/2026/02/03/defi-s-quiet-strength-tvl-holds-as-market-selloff-tests-traders)) 88% Of people think transparency is essential when choosing a crypto platform ([Source](https://www.cointribune.com/en/okx-and-proof-of-reserves-analysis-of-a-transparency-model-that-reassures-crypto-investors/)) 80% Of funds lost in 2024 were due to off-chain attacks ([Source](https://www.halborn.com/reports/top-100-defi-hacks-2025)) ## Infrastructure is where DeFi wins (or loses) Get the latest on Walrus for DeFi, right in your inbox. ## Ready to start building? [Explore the docs](https://docs.wal.app/) Opportunities Built for the systems\ DeFi relies on -------------- ### Auditable execution and strategies Turn execution history and strategy parameters into verifiable records that users and integrators can inspect over time. ### Verifiable market data Publish price and risk data as verifiable objects that smart contracts and partners can reference with confidence. ### Resilient frontends Serve your frontend from Walrus so users can keep trading, even when centralized infrastructure goes down. ## Build on data you can trust [Try Walrus](https://docs.wal.app/) --- # Buy, Stake, and Use WAL: The Walrus Native Token Guide > Get WAL on 18+ exchanges, stake to earn rewards and secure the network, or explore apps built on Walrus. Everything you need to start using WAL in one place. Source: https://walrus.xyz/use/ As markets, payments, and AI agents move onchain, reliable data becomes mission critical. Walrus is built to meet this moment, delivering verifiable data with the speed, availability, and programmability modern systems require. Create a wallet, stake WAL, and help secure the network powering the future of onchain finance. Step 1.\ Create a Wallet --------------- With a Walrus-compatible wallet, you can power data-rich apps, participate in governance, and stake to secure the network. It is your gateway to leveraging Walrus for scalable, onchain storage and accessing the growing ecosystem of applications built on top of it. [A powerful multi-platform wallet built for everyone in the Sui ecosystem](https://slush.app/) [A universal crypto wallet available on app, web, and extension](https://web3.okx.com/) [A simple multichain wallet, reimagined for DeFi and NFTs](https://phantom.app/) [Self-custody multichain wallet delivering first-class Sui ecosystem experience](https://backpack.app/) [Surf Sui + Walrus with this easy, powerful, and seamless wallet](https://surf.tech/) [A multi-chain wallet created to embrace the full potential of blockchain](https://nightly.app/) [Self-custody wallet on Sui, designed for everyone and fully open-sourced](https://suiet.app/) Step 2.\ Buy WAL ------- You can purchase WAL with fiat on supported exchanges or by swapping other cryptocurrencies using supported apps. Choose the method that works best for you. [A crypto exchange offering advanced trading tools to buy and manage crypto](https://www.bybit.com/) [A trusted crypto exchange for buying, selling, and staking digital assets](https://www.kraken.com/) [A global cryptocurrency exchange offering spot and derivatives trading](https://www.bitget.com/) [A leading global crypto exchange, trusted by over 40 million](https://www.kucoin.com/) [A crypto platform to buy, sell, and manage digital currencies with ease](https://crypto.com/) [A digital asset exchange offering trading services across multiple markets](https://upbit.com/) [A simple multichain wallet, reimagined for DeFi and NFTs](https://phantom.com/) [Self-custody multichain wallet delivering first-class Sui ecosystem experience](https://backpack.app/) [A universal crypto wallet available on app, web, and extension](https://www.okx.com/) [A powerful multi-platform wallet built for everyone in the Sui ecosystem](https://slush.app/) [A South Korean crypto exchange offering easy ways to trade digital assets](https://www.bithumb.com/) [A digital asset exchange offering trading services across multiple markets](https://www.mexc.com/) [Uphold is the platform to discover important new tokens early](https://uphold.com/) [A decentralized orderbook exchange for both professional and first-time traders](https://bluefin.io/) [A crypto exchange providing for spot, futures, staking, and more](https://gate.com/) [One of Europe's leading digital asset exchanges](https://bitvavo.com/) [Simplifying trading for all users and assets](https://www.cetus.zone/) [The world's largest cryptocurrency exchange by trading volume, serving 180+ countries](https://www.binance.com/) Step 3.\ Stake your WAL -------------- Stake your WAL tokens to support the Walrus network and earn rewards, or explore applications built on Walrus that use scalable, onchain storage. Walrus is powered by a proof-of-stake network, where WAL token holders can stake to earn rewards in exchange for securing the network. [Find out more](https://blog.walrus.xyz/staking-wal/) To get started, you'll need a compatible wallet and some SUI to cover network fees (learn how to get SUI [here](https://sui.io/get-started)). Then, visit the [official Walrus staking site](https://stake-wal.wal.app/), select a storage node operator from the available list, enter the amount of WAL you wish to stake, and approve the transaction to begin staking. [Explore the Ecosystem](/ecosystem/) --- # WAL Token | Payments, Staking, and Governance on Walrus > WAL is the native token of Walrus, used to pay for storage, secure the network through staking, and vote on protocol governance. Available on 18+ exchanges. Source: https://walrus.xyz/wal-token/ WAL is the native token for Walrus, used for paying fees, securing the network, and participating in governance. [Use WAL](/use/) Foundational\ technology \+ Critical use\ cases \= WAL ## Get WAL Available on more than 18 exchanges worldwide, including: [Discover All Exchanges](/use/#exchanges) The engine that\ powers Walrus ------------- Together, these functions keep costs stable, the network secure, and decision-making transparent and community-driven. ### Payment Use WAL to pay for network fees. The protocol distributes payments over time to nodes and stakers while keeping costs stable, regardless of token price changes. ### Security Secure the network by staking WAL with storage nodes. Nodes earn rewards based on performance, and stakers earn alongside them. ### Governance Walrus governance runs through the WAL token, with voting power based on stake. Nodes vote on system parameters and penalties to keep incentives aligned. WAL\ FAQs ---- ### Token Symbol WAL ### Max Supply 5,000,000,000\ WAL ### Initial Circulating Supply 1,250,000,000\ WAL ### Token Distribution ### Release Schedule ## Ready to use WAL? [Start now](/use/) --- # Walrus Phone Wallpapers > You wouldn't compromise on your data. Don't compromise on your lock screen. Download free Walrus phone wallpapers now. Source: https://walrus.xyz/wallpapers/ Walrus Phone Wallpapers You wouldn’t compromise on your data. Don’t compromise on your lock screen. [Download](/walrus-wallpaper-1.jpg) [Download](/walrus-wallpaper-2.jpg) [Download](/walrus-wallpaper-3.jpg) --- # Walrus for Web3 Builders | Verified Data on Any Chain > Whether you're building on Ethereum, Solana, or another blockchain, Walrus is the Verified Data Platform for builders working with mission-critical data. Source: https://walrus.xyz/web3-devs/ Whether you're building on Ethereum, Solana, or another blockchain, Walrus is the Verified Data Platform for builders working with mission-critical data. [Start building](https://docs.wal.app/) The problem ## Other storage protocols can't keep up Legacy decentralized storage systems fall short when apps demand fast, reliable data and privacy guarantees. Builders are forced to add on centralized infrastructure or DIY encryption, resulting in hybrid stacks that are fragile, expensive, and hard to scale. The solution Your data layer\ should just work ---------------- Walrus gives you the speed, control, and flexibility to build fast, private, and fully onchain applications without workarounds or extra infrastructure. ### Always Available Walrus keeps your critical data online and recoverable by default, so financial systems, AI agents, and data markets can build on a reliable foundation. ### Fast & Verifiable Walrus delivers fast reads and writes with cryptographic verification built in, so you never have to choose between speed and trust. ### Private Walrus lets you define who can access your data and when. With programmable policies and threshold encryption, no infra or custom tooling is required. ### Do even more with the Sui Stack Add onchain access control and verifiable offchain compute to power trusted AI applications. Stay informed ## Data is evolving Stay on top of the game Get Walrus news just for devs, including new features, technical resources, and the latest on what the builder community is up to. ## Start building in minutes Everything you need to go from zero to onchain in just a few clicks. Clear developer guides for setup, storage, and smart contract integration. [Explore Walrus docs](https://docs.wal.app/) Discover how Walrus stacks up next to Filecoin, Arweave, and IPFS. [Compare Walrus](https://docs.wal.app/) Apply to active RFPs and get funding to build with Walrus. [Explore RFPs](/rfp/) ## Trusted by leading teams Helping the most ambitious projects go fully onchain. ### Pudgy Penguins One of the most successful NFT brands of all time. ### Myriad Fast growing multichain prediction market platform. ### Allium The data layer for blockchain analytics. ### OpenGradient Infrastructure for verifiable AI. ### Pipe Network The world's first decentralized, hyperlocal CDN. ### io.net Decentralized computing network for AI. ### One Football The global home for football fans. ### Gata Infra for decentralized AI data and compute.